What the NYU 2022 JAMA Psychiatry proof-of-concept trial actually showed — 83% reduction in heavy drinking days — and where the research goes from here.
Psilocybin for alcohol use disorder is the strongest single controlled result in the psychedelic-addiction field. A double-blind randomized trial run out of NYU found that two dosing sessions embedded in psychotherapy roughly halved the percentage of heavy drinking days versus active placebo — 9.7% vs. 23.6% over 32 weeks, a 13.9-percentage-point difference (p=0.01) that the alcohol-treatment field has not seen from a medication in decades. That result, published in JAMA Psychiatry in 2022, is the reason psilocybin for alcoholism is now being taken seriously by regulators and payers.
The evidence is real, but it is not yet approved care. The trial was a Phase 2 study at two sites with 93 participants. A confirmatory multi-site Phase 3 is enrolling. Nothing about the current legal or clinical landscape lets a person walk into a US clinic and receive psilocybin-assisted treatment for alcohol dependence today. This guide explains what the trial showed, how the therapy was actually delivered, how psilocybin compares with the FDA-approved medications for AUD, and the narrow, honest set of ways someone can access this treatment in 2026.
The 2022 trial by Michael Bogenschutz and colleagues is the pivot point for psilocybin-assisted treatment for alcohol dependence. It was published in JAMA Psychiatry in August 2022 under the title “Percentage of Heavy Drinking Days Following Psilocybin-Assisted Psychotherapy vs Placebo in the Treatment of Adult Patients With Alcohol Use Disorder.”1
The study randomized 93 adults with DSM-IV alcohol dependence to either psilocybin or an active placebo. The active placebo was diphenhydramine (Benadryl) — a sedating antihistamine chosen specifically because inert placebos are trivially unblinded in psychedelic trials. Both arms received the same 12-week manualized psychotherapy: motivational enhancement therapy plus cognitive behavioral elements, delivered by trained therapist dyads.
The two dosing sessions occurred at weeks 4 and 8. The first psilocybin dose was 25 mg per 70 kg. The second was flexible — 25 to 40 mg per 70 kg — based on tolerability. Placebo participants received 50 mg of diphenhydramine at session one and 50 to 100 mg at session two.
The primary outcome was the percentage of heavy drinking days over the 32-week period after the first dose. Heavy drinking was defined as five or more standard drinks in a day for men and four or more for women — the NIAAA cutoff. The psilocybin arm had a mean of 9.7% heavy drinking days. The placebo arm had 23.6%. That is a between-group difference of 13.9 percentage points (95% CI 3.0 to 24.7; p=0.01) and roughly a 59% relative reduction versus placebo. Compared with each arm’s own baseline (both arms started near 52–54% heavy drinking days at intake), the psilocybin arm’s drop was on the order of 83% from baseline — the figure widely quoted in press coverage. The effect was sustained across the follow-up window, not just a transient post-session drop.
Adverse events in the psilocybin arm were mostly mild and expected for the dose — headache, nausea, transient anxiety during the session, elevated blood pressure. There were no serious cardiovascular events attributable to psilocybin and no persistent psychiatric adverse events.
One methodological caveat matters, and competitor summaries usually skip it. Even with diphenhydramine as an active blinding comparator, roughly 93% of psilocybin recipients correctly guessed their assignment at their first session. Diphenhydramine holds the blind better than saline, but it does not hold it well. Expectancy therefore contributes to the observed effect. The multi-site Phase 3 confirmatory trial (NCT05416229) uses a larger sample and additional design features intended to address this. It is enrolling.
The 2022 RCT did not appear from nowhere. It was built on a 2015 open-label pilot by the same team, published in the Journal of Psychopharmacology under the title “Psilocybin-assisted treatment for alcohol dependence: a proof-of-concept study.”2
The pilot enrolled 10 volunteers with active DSM-IV alcohol dependence. Each received motivational enhancement therapy plus one or two psilocybin sessions. There was no control group. Abstinence and heavy drinking days improved sharply after the first psilocybin session and stayed improved at the 36-week follow-up. Baseline drinking predicted little; changes in the intensity of the acute psilocybin experience predicted a lot.
A pilot with 10 people and no control arm cannot prove anything. What it did was justify an FDA-authorized IND, a larger sample, and a real active-placebo control — which is what the 2022 study became.
Psilocybin-assisted treatment for alcohol dependence is not a drug you take — it is a course of psychotherapy in which two of the sessions happen under psilocybin. The 12-week protocol used in the RCT looked like this.
Screening and preparation. Participants completed medical and psychiatric intake, informed consent, and four preparation sessions with the two-therapist dyad. Preparation established rapport, mapped drinking history and motivations, and set intentions for the dosing sessions.
Dosing sessions. Each dose was given in a comfortable clinic room with the two therapists present the entire day. Participants wore eye shades and listened to a curated music playlist designed to support internal focus. Sessions lasted six to eight hours. The therapists did not lead the experience — they held the space, monitored vitals, and were available if the participant wanted to talk.
Integration. Sessions after each dose reviewed what came up, connected it to drinking behavior, and translated insight into action. The overall therapy structure was motivational enhancement therapy plus CBT-based tools for craving and relapse prevention. Psilocybin was the acute catalyst; the therapy framework was where the change was stabilized.
No one has a complete mechanistic explanation. The leading hypotheses are converging rather than settled.
Psilocybin’s primary pharmacology is agonism at the 5-HT2A serotonin receptor in the cortex. That produces the acute subjective effects. In the AUD trials the intensity of the acute mystical-type experience — measured with the Mystical Experience Questionnaire — correlated with drinking-day outcomes. That correlation does not prove the mystical experience causes the effect, but it is consistent with a model in which a high-dose experience reorganizes self-concept and the relationship with alcohol.
A second thread is neuroimaging. Psilocybin acutely disrupts the default mode network — the brain network associated with rumination, self-referential thought, and the entrenched narratives that support addictive patterns. A short window of increased neuroplasticity and cognitive flexibility follows. That window is when integration therapy is designed to operate.
A third thread, still speculative, is direct action on reward circuitry and craving. This has better evidence in psilocybin’s use for tobacco than for alcohol.
AUD already has three FDA-approved medications: naltrexone (oral and long-acting injectable), acamprosate, and disulfiram. These are collectively called medications for alcohol use disorder (MAUD). Anyone considering psilocybin for alcoholism should understand how it sits against the existing options. The table compares the two most-used approved medications with the trial protocol for psilocybin.
| Psilocybin (trial protocol) | Naltrexone | Acamprosate | |
|---|---|---|---|
| Mechanism | 5-HT2A agonist; acute psychedelic experience within a psychotherapy course | Opioid receptor antagonist; blunts alcohol’s rewarding effects | Glutamate/GABA modulator; reduces post-acute withdrawal symptoms |
| Evidence base | 1 Phase 2 RCT (n=93); multi-site Phase 3 enrolling | Dozens of RCTs; Cochrane review supports efficacy | Dozens of RCTs; European trials more favorable than US trials |
| Treatment format | 12-week manualized psychotherapy with two dosing days | Daily pill or monthly injection; typically paired with counseling | Three pills per day; typically paired with counseling |
| Session/dose count | Two psilocybin sessions + ~10 therapy sessions | Ongoing, months to years | Ongoing, months to years |
| Cost (US, 2026) | Not commercially available; trial enrollment is no-cost | Generic oral ~$30/mo; Vivitrol injection ~$1,500/mo | Generic ~$50–$150/mo |
| Access | Clinical trials only | Any prescriber; widely available | Any prescriber; widely available |
When to choose psilocybin. The trial cut heavy drinking days from 23.6% to 9.7% versus active placebo — a 59% relative reduction sustained across 32 weeks — and the effect appears large by AUD-medication standards. But psilocybin-assisted treatment for alcohol dependence is not commercially available. It is not superior enough to established MAUD medications on today’s evidence to justify walking past them — and naltrexone and acamprosate have the advantages of a broad evidence base, straightforward access, and insurance coverage. Psilocybin is worth pursuing for people who have failed or cannot tolerate standard MAUD, who meet criteria for an actively recruiting trial, and who are willing to commit to a 12-week structured psychotherapy program.
There are three access pathways in 2026. Only one is actually treatment for alcohol use disorder.
Clinical trials. This is the only real pathway. The confirmatory multi-site Phase 3 trial of psilocybin-assisted treatment for alcohol dependence is registered on ClinicalTrials.gov and enrolling across several US sites, including NYU, the University of New Mexico, and additional academic centers. Enrollment criteria typically require a current AUD diagnosis, a period of documented sobriety before dosing, and screening for cardiac and psychiatric contraindications. Our clinical trial finder walks through how to search and apply.
Oregon and Colorado service centers. Oregon Measure 109 (implemented 2023) and Colorado Proposition 122 (implementing through 2025–2026) legalized supervised psilocybin use for adults in licensed service centers. These programs are not authorized to treat AUD as a medical condition. A facilitator is not a therapist, service centers are not clinics, and the framework is adult use, not medicine. Some adults with alcohol issues do choose to book a session at a service center and use the experience for their own reasons. That is legal in Oregon and Colorado for adults. It is not the same as receiving psilocybin-assisted treatment for alcohol dependence under a research protocol.
International retreats. Legal psilocybin retreats operate in the Netherlands (using psilocybin truffles), Jamaica, and several other jurisdictions. Almost none run AUD-specific protocols with the manualized psychotherapy structure the trial used. Our best psilocybin retreat centers guide reviews the most-documented operators. Choose a retreat with clear medical screening and understand that the underlying evidence base does not extend to a retreat-format session.
Private psilocybin-assisted therapy is not legally available in US states without a service-center program. Anyone offering it is operating outside the law.
Psilocybin has a strong acute safety profile, but psilocybin-assisted treatment for alcohol dependence is not appropriate for everyone. Trial exclusion criteria are a good baseline for who should not pursue this treatment.
One critical point: psilocybin does not treat alcohol withdrawal. A heavy daily drinker cannot safely stop and start psilocybin. Alcohol withdrawal can cause seizures and death. Any pathway to psilocybin-assisted care for AUD begins with a medically managed detox in a supervised setting. The trial required a period of documented sobriety before the first dose for this reason.
Cost depends entirely on the access pathway.
Clinical trial enrollment is no-cost for accepted participants. The trial pays for screening, therapy, dosing sessions, and follow-up. Travel is sometimes reimbursed. The commitment is 12 weeks of structured therapy plus follow-up assessments over roughly nine months.
Oregon service center sessions currently run $2,000–$4,500 for a supervised adult-use session. That is not AUD-specific care. It is an adult-use session with a trained facilitator. Pricing includes preparation, the dosing session, and integration — but the facilitator does not deliver the manualized psychotherapy the trial used.
International retreats run $3,000–$10,000+ depending on operator and length. Same caveat — not AUD-specific.
Private psilocybin-assisted therapy at a fair market price does not exist in the US in 2026 because it is not legal to offer.
Psilocybin for alcohol use disorder has the single strongest controlled result in the psychedelic-addiction field. The 9.7% vs. 23.6% heavy-drinking-days finding from the 2022 Bogenschutz trial — a 13.9-percentage-point difference (p=0.01), roughly a 59% relative reduction versus placebo and about 83% down from each arm’s own baseline — is large, sustained, and difficult to explain with expectancy alone, even though the trial’s blinding was imperfect. The multi-site confirmatory trial is where the field settles whether that result holds at scale.
For most people with AUD today, the right first step is standard care: a medically managed detox if needed, followed by naltrexone or acamprosate paired with structured therapy. For people who have exhausted those pathways or want to be part of the evidence base, an actively recruiting psilocybin trial is the only real access route. See our clinical trial finder, our legal psilocybin booking checklist for the retreat pathway, and our companion guides on ibogaine for alcoholism, psilocybin for smoking cessation, and the broader psychedelic therapy for addiction field. The parent psilocybin guide covers the underlying pharmacology and legal status in depth.
The strongest evidence to date is Bogenschutz et al.'s 2022 JAMA Psychiatry randomized controlled trial. Two psilocybin sessions embedded in psychotherapy produced a mean of 9.7% heavy drinking days over 32 weeks vs. 23.6% on active placebo (diphenhydramine) — a 13.9-percentage-point between-group difference (p=0.01) and roughly a 59% relative reduction. Effects were sustained; adverse events were mild. A multi-site confirmatory trial is now underway.
In the NYU protocol, patients receive twelve weeks of manualized psychotherapy focused on motivation and coping skills. Two psilocybin sessions (25 mg, then 25–40 mg) are embedded at weeks four and eight, each conducted in a supervised clinical setting with two therapists present, eye shades, and curated music. Therapy sessions before and after each dosing session integrate the experience.
Not as an approved treatment in the US. AUD-specific access is limited to clinical trials — search ClinicalTrials.gov for actively recruiting psilocybin AUD studies at NYU, University of New Mexico, and elsewhere. Oregon (Measure 109) and Colorado (Prop 122) service centers offer supervised psilocybin sessions for adults, but they are not authorized to treat AUD as a medical condition.
The NYU trial screened out participants with severe liver disease, active psychosis, bipolar I disorder, and pregnancy — and required a period of sobriety before dosing. Alcohol withdrawal itself can be dangerous, so any treatment plan should start with a medically managed detox if the patient has been drinking heavily. Psilocybin does not treat withdrawal; it works on the underlying drivers once the patient is stable.
Get psilocybin research & policy updates
New trials, FDA decisions, and legal changes for psilocybin — delivered when they happen.
Suggest a tool, topic, or improvement that would make this site more useful.